There is currently no uniform approach to the follow-up of patients with non-acute cerebellar ataxia. A standardized method for monitoring disease progression is required.

Monitoring the progression of ataxia is essential both to determine the appropriate timing for initiating treatment and to evaluate the effectiveness of interventions.

The Scale for the Assessment and Rating of Ataxia (SARA) is a disease-specific instrument designed to assess the severity of ataxia. However, it is not yet routinely applied in clinical practice. The working group considers the content validity and internal consistency of the instrument to be well established but raises the question whether it sufficiently captures disease progression across different domains.

Summary of Findings

Scale for the Assessment and Rating of Ataxia (SARA) for adults with non-acute cerebellar ataxia

_{^1. Inconsistency (-1 level) :    one  study reports an AUC <0.7}

Description of studies

A total of 12 studies were included in the analysis of the literature. Important study characteristics and results are summarized in table 3. The assessment of the risk of bias is summarized in the risk of bias tables (under the tab ‘Evidence tabellen’).

Reliability

Nine studies reported on reliability of the SARA.

Burk (2009) assessed interrater reliability of the SARA at the same day in 96 individuals with a genetic diagnosis of FRDA. They included patients with a molecular genetic diagnosis of FRDA currently enrolled in the European GeNeMove database.

Kuiper (2014) reported on interrater reliability of the SARA speech subscale in 40 individuals with early onset ataxia. They included children and adults with early onset ataxia described with ‘ataxia’ (as a primary or secondary feature) initiating before their 25^th year of life, as mentioned in the patient records.

Milne (2023) assessed interrater reliability of the SARA in 21 individuals with hereditary cerebellar ataxia between four physiotherapists. They included individuals when: 1) diagnosis was confirmed by genetic testing or a family history of 3 generations affected with cerebellar ataxia 2) age ≥15 years, 3) a level of motor function ranging from difficulty with tandem walking (minimum disability) to unable to walk and requiring minimal assistance with transfers (maximum disability). Exclusion criteria were 1) musculoskeletal injury limiting ability to weight bear, 2) another medical condition affecting mobility, 3) having undergone major orthopedic surgery or received botulinum toxin injections for spasticity in the last 6 months, 4) pregnancy, and 5) significant cognitive impairment.

Schmidt-Hübsch (2006) reported on interrater and test-retest reliability of the SARA in 97 individuals with SCA related ataxia. Inclusion criteria for the patients with SCA were 1) progressive ataxia and 2) positive genetic testing for any SCA mutation or linkage to any SCA locus or clinical evidence of autosomal dominant inheritance.

Schmidt-Hübsch (2010) reported on inter-rater reliability of the SARA in 171 individuals with SCA. Inclusion criteria were positive genetic testing for SCA1, 2, 3, or 6 in combination with progressive, otherwise unexplained ataxia and a complete 1-year follow-up data available for SARA.

Taheri Amin, (2023) compared rating of SARA live and recorded on video in 69 individuals with progressive cerebellar ataxia. Inclusion criteria were (1) progressive cerebellar syndrome and (2) participation in an observational study of the DZNE (Consortium “Sporadic Degenerative Ataxia with Adult Onset,” European Spinocerebellar Ataxia Type 3/Machado-Joseph Disease Initiative, Spinocerebellar Ataxia-Registry, European Friedreich’s Ataxia Consortium for Translational Studies, autosomal recessive cerebellar ataxias).

Viscarra (2025) reported on intra- and interrater reliability of video-rated SARA scores in individuals with spinocerebellar ataxia (SCA) typt 1,2,3 and 6. They performed a secondary data analysis on individuals from the IDEA study. Inclusion criteria were: 1) 18 to 75 years old, 2)genetic conformation of an expanded allele, 3) ability to walk 10 yards for up to 2 minutes without an assistive device and 3) ability to sit or stand unassisted for 30 seconds. Exclusion criteria were: dementia or other comorbid neurological or musculoskeletal conditions.

Wyer (2007) reports inter-rater reliability and test-retest reliability for SARA in individuals with cerebellar ataxia without SCA diagnosis. Inclusion criteria were: 1) progressive or nonprogressive cerebellar ataxia, 2) negative genetic testing for any SCA mutation, and 3) no clinical evidence of autosomal dominant inheritance.

Yabe (2008) reports intra-rater reliability for SARA in 27 individuals with spinocerebellar degeneration.  No in- or exclusion criteria were described.

Responsiveness

Four studies reported on responsiveness of the SARA.

Bourcier (2020) assessed the responsiveness of the SARA in 32 individuals with Autosomal Recessive Spastic Ataxia of Charlevoix-Saguenay (ARSACS). The tested the responsiveness of the SARA in comparison to self-reported Global impression of change of the participants for a 2-year interval. Inclusion criteria were a genetic diagnosis of ARSACS and age between 16 and 65 years old. Exclusion criteria were the presence of other pathologies causing functional limitations, receiving intrathecal baclofen and pregnancy.  

Monte (2017) and Schmidt-Hübsch (2010) assessed the responsiveness of the SARA in relation to the patient’s global impression of change (PGIC) after a 12 month interval.

Winser (2022) reports responsiveness of a SARA balance subscale consisting of the items gait, standing and sitting in relation to the PGIC after 6 and 12 months. Inclusion criteria were: 1) conformed diagnosis of cerebellar ataxia secondary to a genetic or sporadic cause 2) ambulant with or without ambulation device. Exclusion criteria were: 1) bed ridden, 2) unable to perform study instructions and 3) cerebellar ataxia due to acquired lesions.

Table 3. Characteristics of included studies

_{Abbreviations: Ataxia with oculomotor apraxia (AOA) ; BBS- Berg Balance Scale; BI – Barthel Index; CCA  sporadic cortical cerebellar atrophy; CTX - Cerebrotendinous xanthomatosis; DCCA - dominantly  inherited cortical cerebellar atrophy;  Early-onset cerebellar ataxia (EOCA); ICC – intraclass correlation coefficient; MSA - Multiple System Atrophy; MJD - Machado–Joseph  Disease; NR – not reported; PCD- Paraneoplastic cerebellar degeneration; SARA- Scale for the assessment and rating of ataxia; SCA – spinal cerebellar ataxia; SAOA - Sporadic adult-onset ataxia of unknown aetiology; *For further details, see risk of bias tables in the appendix}

Results

1. Reliability (crucial outcome)

The domain reliability refers to the degree to which the measurement is free from measurement error, and it contains the measurement properties internal consistency, reliability, and measurement error. Reliability was reported by:

1.1 Inter-rater reliability

Table 4 shows the inter-rater reliability reported by xx studies. Inter-rater reliability was high for all studies.

Table 4. Inter-rater reliability

Single item inter-rater reliability

Schmidt-Hübsch (2006) reports an inter-rater reliability (ICC) >0.80 for all single items except for item 6 (nose-finger test) and item 8 (heel-shin slide). No exact values or confidence intervals are reported.

Vizcarra (2025) reports an inter-rater reliability for individual items of: ICC  0.87 to 0.96.

Weyer (2007) reports an inter-rater reliability (ICC) for individual items scores >0.80 except for item 6 (nose-finger) with an ICC of 0.793 (95% CI not reported).

Kuiper (2014) reports an ICC of 0.854 (95%CI not reported) for the SARA speech subscore.

1.2 Intra-rater reliability/test-retest reliability

Table 5. Test-retest reliability

_{^* This result is based on a subset of 15 intra-rater assessments. ^** analysis on patients that report a stable impression (n=43).  † Subset of 18 patients.}

Single item reliability

Milne (2023) reports single item ICCs ranging from 0.34 (Item 7; Fast alternating hand movements) to 0.94 (Item 1 and 2; Gait and Stance).

Vizcarra (2025) reports an ICC for intra-rater reliability for individual items of: 0.82 (sitting) to 0.96 (fast alternating hand movement).

Taheri Amin (2023) reports an ICC for live vs video rating between 0.82 (nose-finger test) to 0.98 (gait) for single item analysis.

Weyer (2007) report that test-retest reliability for subitems was lower. They report the lowest scores for item 3 (sitting; ICC: -0.133) and item 6 (nose-finger test; ICC: 0.498). 

Yabe (2008) reports an ICC of individual item scores >0.80 except for item 5 (finger chase) with an ICC of 0.712 and item 7 (fast alternating hand movements) with an ICC of 0.742.

Bland-Altman analyses video compared to live rating

In addition to table 4, Taheri Amin (2023) reports a Bland Altman bias of 0.09 ± 2.0 with 95% limits of agreement (LOA) [-3.85 to 4.03] for video rating compared to live rating. The direction of this difference is not clear.

Vizcarra (2025) reports a Bland Altman bias of 0.47 ±1.24 with 95% LOA [-1.96 to 2.89] indicating that video-ratings overestimated live ratings by 0.47 points on average.

2. Responsiveness (crucial outcome)

Three studies defined responsiveness as the ability to discriminate between patients self-reported change in function with the Patients’ Global Impression of Change (PGIC) (Monte, 2017; Schmidt-Hübsch, 2010; Winser, 2022). Patients were dichotomized between those who report a change in function and those who do not.

Monte (2017) reports an ROC (AUC) of 0.81 (95%CI: 0.58 to 1.0] for the ability of the SARA to discriminate between patients who report a change in function with the PGIC and those who do within not in a 12-month interval. They considered this sufficient responsiveness according to their hypothesis (AUC > 0.7). The best cutoff value for SARA was 1.25 points, with 81.3% of sensitivity and 87.5%  specificity.

Schmidt-Hübsch (2010) reports an AUC of 0.613 for the ability of the SARA to discriminate between patients who report a change in function with the PGIC and those who do not in a 12-month interval. This was insufficient responsiveness according to their hypothesis (AUC>0.7). No cutoff points were reported.

Winser (2022) reports an ROC (AUC) of 0.81 at 6 months and of 0.76 at 12 months for the SARA balance score to discriminate between patients who report a change in function with the PGIC (n = 24) and those who do not (n=12). They also report ROC (AUC) for the complete SARA: 0.86 at 6 months and 0.74 in 12 months. No cutoff points were reported.

Bourcier (2020) defined responsiveness as the ability to discriminate between patients self-reported progression of the disease with sub question on balance and gait in a Global Impression Rating of Change questionnaire. Patients were dichotomized by those with ‘no change’ (score of 3- remained stable) and those with ‘progression’.

They did not find a relation between self-reported disease progression and SARA score within 2 years. No absolute SARA scores are reported for both groups.

 “The responsiveness to change was not significant, and the interpretation of change score increased by 1.9 ± 2.5 falling below the minimal detectable change threshold of 3.06.” This study is not assessed via the GRADE methodology because it is unclear whether the results are in accordance with the hypothesis.

3. Feasibility (important outcome)

Schmidt-Hubz (2006) reports a mean time to administer of 14.2 ± 7.5 minutes (range 5 to 40 minutes). Yabe (2008) reports a mean time to examine of 4.3 ± 0.63 minutes.

What are the clinimetric properties of the SARA scale for individuals with non-acute cerebellar ataxia?

Table 1. PICO

Relevant outcome measures

The guideline panel considered responsiveness and reliability (test-retest and inter-rater) as a critical outcome measure for decision making; and feasibility as an important outcome measure for decision making.

The measurement properties were defined following the taxonomy of the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) (Mokkink, 2010; figure 1, table 2).

_{Figure 1. COSMIN taxonomy of measurement properties}

The guideline panel defined the values listed in table 2 as sufficient measurement properties (according to COSMIN; see Prinsen, 2018). Responsiveness was further defined as the ability of the SARA to determine a change in ataxia severity over time.

The COSMIN Risk of Bias tool was used to assess the quality of included studies for each measurement property (Mokkink, 2024). The worst-score-counts method was used to determine the risk of bias, meaning that the lowest rating given in a box determines the final rating, i.e., the quality of the study.

Table 2. COSMIN definitions of measurement properties (Mokkink, 2010) and COSMIN criteria for sufficient measurement properties (Prinsen, 2018)

_{^a To rate the quality of the summary score, the factor structures should be equal across studies.}

_{^b Unidimensionality refers to a factor analysis per subscale, while structural validity refers to a factor analysis of a (multidimensional) patient-reported outcome measure.}

_{^c The results of all studies should be taken together and it should then be decided if 75% of the results are in accordance with the hypotheses.}

_{^d This evidence may come from different studies.}

In addition, feasibility was described by looking at reported time to administer, costs, personnel, and percentage of individuals for whom the test could be performed. This outcome measures described only and not assessed via GRADE methods.

Search and select (Methods)

A systematic literature search was performed by a medical information specialist using the following bibliographic databases: Embase.com and Ovid/Medline all. Both databases were searched from 2005 to March 7^th, 2025 for systematic reviews, RCTs and observational studies. Systematic searches were completed using a combination of controlled vocabulary and natural language keywords. The overall search strategy was derived from the following primary search concepts: (1) cerebellar ataxia; (2) Scale for Assessment and Rating of Ataxia (SARA); (3) clinimetrics. Duplicates were removed using EndNote software. After deduplication a total of 470 records were imported for title/abstract screening. Initially, 44 studies were selected based on title and abstract screening. After reading the full text, 32 studies were excluded (see the exclusion table under the tab ‘Evidence tabellen’), and 12 studies were included.

1. Berger M, Garcia-Moreno H, Ferreira M, Hubener-Schmid J, Schaprian T, Wegner P, Elter T, Teichmann KM, Santana MM, Grobe-Einsler M, Oender D, Koyak BSC, Bernsen S, Pereira de Almeida L, Silva P, Ribeiro JA, Cunha I, Gonzalez-Robles C, Khan S, Heslegrave A, Zetterberg H, Lima M, Raposo M, Ferreira AF, Vasconcelos J, van de Warrenburg BP, van Gaalen J, van Prooije TH, de Vries J, Schols L, Riess O, Synofzik M, Timmann D, Thieme A, Erdlenbruch F, Infante J, Pelayo-Negro AL, Manrique L, Reetz K, Dogan I, Oz G, Joers JM, Bushara K, Onyike C, Povazan M, Jacobi H, Schmahmann JD, Ratai EM, Schmid M, Giunti P, Klockgether T, Faber J. Progression of biological markers in spinocerebellar ataxia type 3: longitudinal analysis of prospective data from the ESMI cohort. Lancet Reg Health Eur. 2025 Jul 3;55:101339. doi: 10.1016/j.lanepe.2025.101339. PMID: 40678042; PMCID: PMC12270660.
2. Bourcier D, Bélanger M, Côté I, Brais B, Synofzik M, Brisson JD, Rodrigue X, Gagnon MM, Mathieu J, Gagnon C. Documenting the psychometric properties of the scale for the assessment and rating of ataxia to advance trial readiness of Autosomal Recessive Spastic Ataxia of Charlevoix-Saguenay. J Neurol Sci. 2020 Oct 15;417:117050. doi: 10.1016/j.jns.2020.117050. Epub 2020 Jul 18. PMID: 32736199.
3. Bürk K, Mälzig U, Wolf S, Heck S, Dimitriadis K, Schmitz-Hübsch T, Hering S, Lindig TM, Haug V, Timmann D, Degen I, Kruse B, Dörr JM, Ratzka S, Ivo A, Schöls L, Boesch S, Klockgether T, Klopstock T, Schulz JB. Comparison of three clinical rating scales in Friedreich ataxia (FRDA). Mov Disord. 2009 Sep 15;24(12):1779-84. doi: 10.1002/mds.22660. PMID: 19562766.
4. Kuiper MJ, Brandsma R, Lawerman TF, Lunsing RJ, Keegstra AL, Burger H, De Koning TJ, Tijssen MAJ, Sival DA. Assessment of speech in early-onset ataxia: a pilot study. Dev Med Child Neurol. 2014 Dec;56(12):1202-1206. doi: 10.1111/dmcn.12517. Epub 2014 Jun 18. PMID: 24942085.
5. Milne SC, Roberts M, Ross HL, Robinson A, Grove K, Modderman G, Williams S, Chua J, Grootendorst AC, Massey L, Szmulewicz DJ, Delatycki MB, Corben LA. Interrater Reliability of the Scale for the Assessment and Rating of Ataxia, Berg Balance Scale, and Functional Independence Measure Motor Domain in Individuals With Hereditary Cerebellar Ataxia. Arch Phys Med Rehabil. 2023 Oct;104(10):1646-1651. doi: 10.1016/j.apmr.2023.05.003. Epub 2023 Jun 1. PMID: 37268274.
6. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, Bouter LM, de Vet HC. The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol. 2010 Jul;63(7):737-45. doi: 10.1016/j.jclinepi.2010.02.006. PMID: 20494804.
7. Mokkink LB, Elsman E, Terwee CB. The COSMIN guideline for systematic reviews of Patient-Reported Outcome Measures (PROMs). Qual Life Res. 2024; doi: 10.1007/s11136-024-03761-6.
8. Monte TL, Reckziegel ER, Augustin MC, Silva ASP, Locks-Coelho LD, Barsottini O, Pedroso JL, Vargas FR, Saraiva-Pereira ML, Leotti VB, Jardim LB; Rede Neurogenética. NESSCA Validation and Responsiveness of Several Rating Scales in Spinocerebellar Ataxia Type 2. Cerebellum. 2017 Aug;16(4):852-858. doi: 10.1007/s12311-017-0855-8. PMID: 28456900.
9. Moulaire P, Poulet PE, Petit E, Klockgether T, Durr A, Ashisawa T, Tezenas du Montcel S; READISCA Consortium. Temporal Dynamics of the Scale for the Assessment and Rating of Ataxia in Spinocerebellar Ataxias. Mov Disord. 2023 Jan;38(1):35-44. doi: 10.1002/mds.29255. Epub 2022 Oct 23. PMID: 36273394; PMCID: PMC9851985.
10. Prinsen CAC, Mokkink LB, Bouter LM, Alonso J, Patrick DL, de Vet HCW, Terwee CB. COSMIN guideline for systematic reviews of patient-reported outcome measures. Qual Life Res. 2018 May;27(5):1147-1157. doi: 10.1007/s11136-018-1798-3. Epub 2018 Feb 12. PMID: 29435801; PMCID: PMC5891568.
11. Schmitz-Hübsch T, du Montcel ST, Baliko L, Berciano J, Boesch S, Depondt C, Giunti P, Globas C, Infante J, Kang JS, Kremer B, Mariotti C, Melegh B, Pandolfo M, Rakowicz M, Ribai P, Rola R, Schöls L, Szymanski S, van de Warrenburg BP, Dürr A, Klockgether T, Fancellu R. Scale for the assessment and rating of ataxia: development of a new clinical scale. Neurology. 2006 Jun 13;66(11):1717-20. doi: 10.1212/01.wnl.0000219042.60538.92. Erratum in: Neurology. 2006 Jul 25;67(2):299. Fancellu, Roberto [added]. PMID: 16769946.
12. Schmitz-Hübsch T, Fimmers R, Rakowicz M, Rola R, Zdzienicka E, Fancellu R, Mariotti C, Linnemann C, Schöls L, Timmann D, Filla A, Salvatore E, Infante J, Giunti P, Labrum R, Kremer B, van de Warrenburg BP, Baliko L, Melegh B, Depondt C, Schulz J, du Montcel ST, Klockgether T. Responsiveness of different rating instruments in spinocerebellar ataxia patients. Neurology. 2010 Feb 23;74(8):678-84. doi: 10.1212/WNL.0b013e3181d1a6c9. PMID: 20177122.
13. Taheri Amin A, Faber J, Önder D, Kimmich O, Synofzik M, Ashizawa T, Klockgether T, Grobe-Einsler M. Comparison of Live and Remote Video Ratings of the Scale for Assessment and Rating of Ataxia. Mov Disord Clin Pract. 2023 Aug 7;10(9):1404-1407. doi: 10.1002/mdc3.13843. PMID: 37772290; PMCID: PMC10525045.
14. Vizcarra JA, Casey HL, Hamedani AG, Gomez CM. Reliability of remote video ratings of the scale for assessment and rating of ataxia. Parkinsonism Relat Disord. 2025 Mar;132:107278. doi: 10.1016/j.parkreldis.2025.107278. Epub 2025 Jan 13. PMID: 39832443.
15. Weyer A, Abele M, Schmitz-Hübsch T, Schoch B, Frings M, Timmann D, Klockgether T. Reliability and validity of the scale for the assessment and rating of ataxia: a study in 64 ataxia patients. Mov Disord. 2007 Aug 15;22(11):1633-7. doi: 10.1002/mds.21544. PMID: 17516493.
16. Wilke C, Pellerin D, Mengel D, Traschütz A, Danzi MC, Dicaire MJ, Neumann M, Lerche H, Bender B, Houlden H; RFC1 study group; Züchner S, Schöls L, Brais B, Synofzik M. GAA-FGF14 ataxia (SCA27B): phenotypic profile, natural history progression and 4-aminopyridine treatment response. Brain. 2023 Oct 3;146(10):4144-4157. doi: 10.1093/brain/awad157. PMID: 37165652.
17. Winser SJ, Chan AYY, Chung R, Whitney S, Kannan P. Validity of balance measures in cerebellar ataxia: A prospective study with 12-month follow-up. PM R. 2023 Jun;15(6):742-750. doi: 10.1002/pmrj.12826. Epub 2022 Jul 8. PMID: 35474301.
18. Yabe I, Matsushima M, Soma H, Basri R, Sasaki H. Usefulness of the Scale for Assessment and Rating of Ataxia (SARA). J Neurol Sci. 2008 Mar 15;266(1-2):164-6. doi: 10.1016/j.jns.2007.09.021. Epub 2007 Oct 22. PMID: 17950753.